Please describe your situation in more detail.
How did you generate the set of candidate items? Were they from
pre-existing scales that represent similar constructs? A newly
generated set of items from dissimilar constructs?
If you did some kind of factor analysis, what kind? what stopping rules
did you use?
How did you develop the scoring key? Pre-existing? Unit weights? other
weights? Did you use only cleanly loading items?
Are you trying to examine divergent validity? Create proxy measure for
the set of constructs?
Are you talking about variance only within the set of scales? Or are
you talking about their unique addition to predicting/explaining/fitting
to some criterion/dependent variable?
Art Kendall
Social Research Consultants
Hannah State-Davey wrote:
> All,
>
> I am developing a measure which has a number of subscales. Examination of
> the inter-scale correlations revealed that they are moderately to highly
> correlated. I would therefore like to determine the amount of unique
> variance of each subscale - i.e. their level of independence from the rest
> of the subscales.
>
> Can anyone advise on how I do this in SPSS?
>
> Many thanks in advance.
>
> Hannah
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
>
[hidden email] (not to SPSSX-L), with no body text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the command
> INFO REFCARD
>
>
>
=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Art Kendall
Social Research Consultants
Locked
